Improved image matching method based on cursory search and detail-oriented correction with extension window

Abstract

In order to achieve the fast and accurate image matching, gray matching algorithm and SIFT feature matching algorithm are combined, and an approach to the cursory search and detail-oriented correction with extension window is proposed. The cursory search is achieved by using new adaptive optimal guidance artificial bee colony algorithm (AOGABC) instead of ergodicity of the traditional gray matching algorithm. The gray correlation degree with statistical properties serves as the fitness function of the artificial colony algorithm (ABC). The extensional image window has built after cutting image according to the extension rules in extension window, detail-oriented correction accurately matches image by using the SIFT algorithm. The experiments verify that the matching method not only realizes rapidity because of performance of artificial bee colony algorithm and gray relational grade in the cursory search, but also achieves matching accuracy resulted from the combination of SIFT algorithm and extension window in this paper. By comparing the effects of different algorithms in the typical image, the results show that the purpose of the exact match is achieved.